The purpose of this project is to investigate the nature of knowledge representations in children with Specific Language Impairments (SLI). Processing based accounts of SLI have focused primarily on the extent to which limitations in processing capacity account for the language impairments seen in these children. In particular, these accounts have focused primarily on the extent to which slower rates of processing or less efficient access to information leads to capacity limitations in SLI. Within these accounts, little emphasis has been placed on the nature of the information being manipulated. Recently, however, it has been suggested that mental representations in children with SLI may differ from those of typically developing children. Specifically, it has been suggested that children with SLI may be working with mental representations that are more concrete and grounded in nature and thus more cumbersome and inefficient to manipulate and store as compared to more abstract representations used by their normal language peers. The purpose of this project is to investigate the extent to which children with SLI are relying on representations that are more grounded as compared to NL controls. A total of 148 children will participate in this 5-year project, including 74 children with SLI, 7-9 years of age, 34 normal language controls matched on mental age (MA), nonverbal l.Q., gender, and SES, and 40 younger NL controls matched on gender, SES, and quantitative knowledge (QK). A total of 6 cognitive problem solving studies are proposed. Studies 1-4 will examine the impact of grounded versus abstract problem representations on the speed and accuracy of problem solutions for children with SLI (N = 34) and MA controls (N = 34). Studies 5 and 6 will examine the nature and rate of change of knowledge representations for a group of children with SLI (N = 40) and a group of younger QM controls (N = 40). Study 5 will examine children's spontaneous problem solutions for mathematical problems and Study 6 will use a microgenetic design to examine the nature and rate of shifts in representations of mathematical knowledge for children with SLI (N = 40) and QM controls (N = 40). The outcomes of this project will serve to advance our understanding of the nature of processing deficits in children with SLI, and provide valuable insights into the development of effective models of intervention.